Our central hypothesis is that a patient's therapy selects intrinsically resistant cells which share genetic changes that confer a selective advantage, enabling such cells to repopulate the tumor mass. This hypothesis is predicated on our previous studies in which tumors selected with BCNU in vitro demonstrated a non-random karyotypic deviation consisting of polyploidy of chromosomes 7 and 22 (7/22). This cytogenetic abnormality was detected in approximately 30% of the tumors with standard cytogenetic techniques. Seven patients who were known to have this subpopulation required second surgical procedures permitting us to compare the cells of the primary tumor to tumor that had received adjuvant therapy. The dominant population in all seven patients demonstrated the polyploidy of chromosomes 7/22. Biochemical and molecular characterization demonstrated that these cells aberrantly expressed platelet-derived growth factor (PDGF). Preliminary evidence suggests that these cells are also resistant to irradiation and cisplatin. These results suggest that this non-random karyotypic deviation defines a subset of glioma cells that may have additional phenotypic similarities related to growth and/or cellular resistance. To test the hypothesis that such cells will share other common phenotypes we now propose to continue the characterization of additional primary/recurrent tumor pairs to determine if cells with polyploidy of chromosomes 7/22 aberrantly express one or more growth factor/growth factor receptor abnormalities, specifically focusing on PDGF (mapped to chromosomes 7 and 22) and epidermal growth factor receptor (EGFR) (mapped to chromosome 7), a receptor which is aberrant in about 30% of gliomas. We plan to characterize these genes using standard and molecular cytogenetics, Southern blot hybridization, reverse transcription followed by the polymerase chain reaction (RT-PCR) and western blot analyses. Further, we propose to determine if one or more genes specifically related to BCNU resistance such as methylguanine methyltransferase and/or glutathione-S-transferase are abnormally expressed in these cells. We will compare the results of our in vitro studies with the expression of these same genes in the patient's tissues utilizing fluorescent in situ hybridization (FISH), mRNA in situ hybridization and immunohistochemistry. These studies will provide an in vivo correlate to our in vitro analyses. Our long range goal is to identify a panel of characteristics within the architecture of the brain tumor that may provide the targets for the development of new drugs and/or new therapeutic strategies.